Mission Impossible? Blaming Primary Care Providers for Not Identifying the Unidentifiable

In this issue of Pediatrics, Nelson et al report that at 2 years of age, parental, social, and economic factors (including parental education levels below a bachelor’s degree, little or no shared reading at home, food insecurity, and fair/poor parental health), rather than direct developmental evaluation, best predict future low academic achievement scores and high problem behaviors at kindergarten entry.1 Previously, the predictive validity of the Bayley Scales of Infant Development—II performed at 20 months of age to predict IQ scores at 8 years of age has been shown to be only 37%.2 If such comprehensive developmental evaluation using gold-standard assessment measures perform so poorly in predicting learning and behavioral problems at school entry, is it time to stop blaming primary pediatric health care providers for not identifying these children?More than half a century ago, several well-controlled studies demonstrated weaknesses in pediatricians’ ability to assess children developmentally.3,4 Furthermore, in 1987, it was reported that only 28.7% of children who required special educational services in elementary school were identified before they reached school at 5 years of age.5 Subsequently, a perception has developed that primary pediatric health care providers, who are the professionals most responsible for identifying children before school entry, are to blame for missing the vast majority of children with developmental and behavioral disabilities before they get to school. This has led to mandated use of parent-completed developmental and behavioral screening questionnaires and an erosion of confidence in primary pediatric health care providers’ clinical skills. However, these studies antedate changes in pediatric residency education that required rotations in developmental-behavioral pediatrics and subspecialty-trained developmental-behavioral pediatric faculty in all pediatric residency training programs. The only study completed in this millennium that directly compared pediatricians’ developmental assessments with those performed by child psychologists showed that whereas pediatricians failed to identify 24% of children with mild neurodevelopmental impairments, child psychologists failed to identify 39% of these children.6Although lower morbidity neurodevelopmental disorders, such as learning disabilities, attention-deficit/hyperactivity disorder (ADHD), and developmental coordination disorders, are far more prevalent than higher morbidity disabilities, such as intellectual disabilities, autism spectrum disorder, and cerebral palsy, it is a basic neurodevelopmental tenet that the milder the disability, the older the age that it can be reliably identified. The developmental trajectory of most children with intellectual disability, autism spectrum disorder, and cerebral palsy should be identified as different from typical by 3 years of age, but the developmental trajectory of children with learning disabilities, ADHD, and motor incoordination may not be distinguishable from those with typical development until faced with the academic and social demands of school. Certainly, autism cannot be identified at 2 months, dyslexia cannot be identified at 2 years, and (given that some degree of inattentiveness and impulsiveness characterize the behavior of toddlers and preschoolers), we will leave it to the reader to specify a reasonable age for the identification of ADHD. Thus, in the early intervention (birth to 3 years) age group, only the less prevalent, higher morbidity disabilities are capable of being reliably identified. The Centers for Disease Control and Prevention reported that the 1997–2008 overall prevalence rates of intellectual disability, autism, and cerebral palsy were 0.71%, 0.47%, and 0.39%, respectively, with the 2006–2008 prevalence of autism increasing to 0.74% and intellectual disability decreasing to 0.67%.7 Given these prevalence rates for developmental-behavioral disorders that are truly identifiable in the birth to 3 years age group, it is reassuring that in 2015, 2.95% of children nationwide were receiving Part C early intervention services.8 Thus, these data suggest that the number of children receiving early intervention services exceeds the number of children with identifiable developmental-behavioral disabilities. Rather than missing the majority of these children, these data suggest that primary pediatric health care providers may actually be doing very well in identifying those children with developmental-behavioral disabilities who are able to be reliably identified in the birth to 3 year age range. Certainly the majority of children who require special education are not identified before school entry. However, the majority of children who receive special education (68%) qualify for special education services as a result of language-learning disabilities or other health impairments (which include children with ADHD) that cannot always be reliably identified before school age, even with comprehensive developmental evaluations using gold standard assessment measures.9Thus, it may be time to acknowledge primary pediatric health care providers for often identifying those children who can be identified and stop blaming them for not identifying the unidentifiable. Instead, the article by Nelson et al1 suggests the use of psychosocial screening as an alternative to continued unrealistic expectations for primary pediatric health care providers to identify children at risk for the higher prevalence, lower morbidity learning and behavioral problems that are not reliably identifiable at younger ages.